Título: Algoritmo Evolutivo Multi-objetivo de Tabelas para Seleção de Variáveis em Calibração Multivariada
Autores: Jorge, Carlos Antônio Campos; Lima, Telma Woerle de; Ribeiro, Lucas de Almeida; Filho, Arlindo R. G.; Coelho, Clarimar José; Soares, Anderson da Silva; Delbem, Alexandre C. B.
Resumo: This paper proposes a new multiobjective evolutionary algorithm. In particular, this algorithm makes use of a structure of subsets stored in a data structure called table in which the best individuals from each target are considered preserved. This approach is compared in this paper with traditional evolutionary algorithms monoobjective. A case study is presented a problem multivariate calibration which involves the prediction of protein concentration in samples of whole wheat from spectrophotometric measurements wave. The results show that the proposed formulation has a smaller prediction error as compared to the formulation monoobjective. Additionally, it was observed that the templates are obtained from the proposed formulation using a lower number of variables. Finally, a study of sensitivity to noise in the models obtained by multi-objective formulation showed better result.
Código DOI: 10.21528/CBIC2013-236
Artigo em pdf: bricsccicbic2013_submission_236.pdf
Arquivo BibTex: bricsccicbic2013_submission_236.bib